首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于BP网络与Mahalanobis距离的图像识别技术(英文)
引用本文:梁忠伟,叶邦彦,刘晓初.基于BP网络与Mahalanobis距离的图像识别技术(英文)[J].科学技术与工程,2009,9(11).
作者姓名:梁忠伟  叶邦彦  刘晓初
作者单位:1. 广州大学,机械与电气工程学院,广州,510006
2. 华南理工大学机械与汽车工程学院,广州,510640
摘    要:提出基于BP人工神经网络与Mahalanobis距离的图像识别技术.在对IC芯片图像进行二维小波分析后,通过Mahalanobis距离在图像的低频小波系数矩阵上进行目标图元的区域搜索与位置标定,而后根据待定目标图元像素分布特征建立特征参数向量,建立BP网络对目标图元进行模式识别.实验证明此方法能够有效地对IC图像进行目标图元的位置标定与模式识别,获得准确的IC标定识别结果.

关 键 词:Mahalanobis距离  BP网络  图像识别  芯片检测

Image Recognizing Technology Based on BP Network and Mahalanobis Distance
LIANG Zhong-wei,YE Bang-yan,LIU Xiao-chu.Image Recognizing Technology Based on BP Network and Mahalanobis Distance[J].Science Technology and Engineering,2009,9(11).
Authors:LIANG Zhong-wei  YE Bang-yan  LIU Xiao-chu
Institution:School of Mechanical and Electrical Engineering;Guangzhou University.Guangzhou 510006 P.R.China;School of Mechanical and Automobile Engineering;South China University of Technology1.Guangzhou 510640 P.R.China
Abstract:For purpose of IC chip detection,a new image recognizing and calibrating technology based on Mahalanobis distance and BP artificial neural network is proposed.After IC chip's micro-image has been wavelet analyzed,Mahalanobis distance is calculated and used to implement IC object's region searching and position calibrating,then its feature parameters are calculated,BP network is built up for object's pattern recognition.Experiment proves that this method can achieve IC image's position calibration and patter...
Keywords:Mahalanobis distance BP artificial neural networks Image recognition IC chip detection
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号